Development and Applications of a Four-Channel Enzyme Thermistor System for Bioprocess Control

De elopmen and Applica ions o a Fou -Channel Enzyme The mis o Sys em o Biop ocess Con ol H.-G. Hundeck, U. Hubne , A. Lübbe , T. Schepe , J. Schmid , M. Weiß Ins i u ü Technische Chemie, Uni e si ä Hanno e , Callins . 3, D-3000 Hanno e 1 F. Schube Zen alins i u ü Mik obiologie, Robe -Rössle-S ., D-1115 Be lin-Buch Summa y A la ge numbe o pape s on biosenso s ha e been published in he las ew yea s. Howe e , ew o hese sophis ica ed analysis sys ems ha e been used o moni o eal biop ocesses. In his pape , a newly de eloped ou -channel enzyme he mis o sys em and i s applica ion o bio echnological p ocess moni o ing is p esen ed and discussed. Di e en suga s we e de ec ed simul aneously and on- line du ing he cul i a ion o Spodop e a ugipe da and Bacillus licheni o mis in echnical media. Immobilized enzymes and en apped mic oo ganisms we e used as biological compound in his biosenso . In addi ion, enan ioselec i e analysis was pe o med by wo enzyme eac ions. Fo example, he de ec ion o D,L- acema es o aminoacid es e s was p esen ed in an aqueous sys em. Fu he mo e he possibili y o using his de ec ion sys em in o ganic sol en s was shown. In oduc ion On-line moni o ing echniques a e impo an o op imal biop ocess con ol and au oma ion. This ype o measu emen places special demands on he ypes o senso s in ol ed (e.g., long e m- s abili y, analy ical accu acy, and au oma ed analy ical p ocedu e). A majo ocus o bio echnological analysis is he cul i a ion medium, since subs a es and p oduc s a e dissol ed in i . The en i onmen o he cells o be cul u ed in luences he ac i i y and he s a e o he biomass and hus o he whole biop ocess. O en, he analysis o he cul i a ion medium is pe o med by complex o -line me hods. Biosenso s p o ide he possibili y o measu ing such com- pounds wi h on-line sys ems, and g ea ly inc ease he powe o au oma ed biop ocess moni o ing and con ol. The li e a u e on biosenso s is o e whelming (1-6), bu he e is s ill a lack o applica ions in eal indus ial p ocesses (7). The ou -channel enzyme he mis o sys em is un as a s and-alone de ice, con olled by a compu e . The simul aneous analysis o di e en mono- and disaccha ides such as glucose, mal ose, suc ose, and lac ose wi h his e sion was simply 322 H.-G. Hundeck e al. pe o med a cul i a ion p ocesses o e pe iods o up o 300 hou s and can hus be used as a basis o p ocess op imiza ion. Ma e ial and Me hods Ma e ials Glucose oxidase (EC 1.1.3.4) and ca alase (EC 1.11.1.6) we e used o glucose analysis. In e ase (EC 3.2.1.26) was used o suc ose analysis, and a-glucosidase (EC 3.2.1.20) was employed o mal ose analysis. The enzymes we e co alen ly bound on oxi ane ac ylic suppo s (VA Epoxy Biosyn h, Riedel de Haen AG, W. Ge many) (8). This kind o immobiliza ion is mo e e icien han he usual CPG immobiliza ion (9). The o ganisms used in he senso (e.g., cells o Saccha omyces ce e isiae) we e immobilized in calcium algina e (10). The Enzyme The mis o The analy ical sys em is based on a simple isope ibol low calo ime e , in which he hea p oduced is measu ed as a empe a u e change by he mis o s (11). The empe a u e esolu ion is abou 10-5 K. A schema ic diag am o an enzyme he mis o is shown in Figu e 1. e e ence enzyme column gold ube A— LEA ELL LE „a a Y A A / ; J Y 5 YZ ’ 7 4} 7 | 7 7 he mis o i) ] a hea exchange 4 A A y A aluminium cylinde 5 A V 1 [A insula ion VLAZZLZ. LAL LZ] a © 1 4 sample injec ion bu e Ki ee bu e Fig. 1 Schema ic diag am o an enzyme he mis o . A Fou -Channel Enzyme The mis o Sys em o Biop ocess Con ol 323 Bu e is pumped con inuously h ough he he mos a ed aluminium cylinde . Be o e en e ing he columns, he ca ie low passes h ough a hea exchange coil. The empe a u e a he inle and ou le o he column is egis e ed by he mis o s placed on gold ubes. One o he columns con ains he immobilized biological ma e ial (e.g., enzymes), while he o he column is illed wi h inac i a ed esin ma e ial. In his e e ence column, nonspeci ic hea e ec s a e measu ed. Bu e is pumped con inuously h ough he analysis sys em. A de ined sample olume is injec ed by an au oma ed al e in o he bu e s eam acco ding o he p inciples o low injec ion analysis (12). The ca ie low anspo s he sample h ough he columns; he eac ion occu s in he enzyme column. The hea p oduced he e is measu ed by he he mis o s. analog ampli ie 68000 mic op ocesso Fig. 2 Compu e con ol sys em o he ou -channel enzyme he mis o . A ou -channel e sion o an enzyme he mis o was designed and buil in he Ins i u ü Technische Chemie as a low calo ime e , based on he expe iences wi h Lund- ype enzyme he mis o s (13-14). I can analyze ou di e en subs ances simul- aneously. A e changing an enzyme column, only 15 minu es a e necessa y be o e he analysis can be con inued. Special columns ha e been designed o he de ec ion o dissol ed enzyme ac i i ies. In 324 H.-G. Hundeck e al. hese columns, subs a e and enzymes a e mixed oge he . The hea measu ed is p opo ional o he enzyme ac i i y.In eg a ed p e- he mos a ing c ea es a e y s able empe a u e a he measu emen cen e . A mic op ocesso con ol uni is u ilized o he signal de ec ion (Figu e 2) . The use can de ine he analysis channels wi h he so wa e. He can choose a e e ence column, a measu emen be ween wo he mis o s o he same column, o o he special measu emen applica ions. The con ol sys em checks he measu emen accu acy, calib a ion, and da a analysis and p ocessing. The 68000 mic o- p ocesso sys em wi h a mul i- asking ope a ing sys em allows he implemen a ion o con o! algo i hms o sophis ica ed p ocess con ol; e.g., he egula o con ols he subs a e eed o biop ocesses ia he pump low. A special il a ion sampling p obe was in eg a ed in o he whole analy ical sys em. A cell- ee sample can be wi hd awn om he e men o con inuously o injec ion in o he enzyme he mis o . Dynamic o s a ic dilu ion s eps a e ini ia ed by he compu e i necessa y. The sample is injec ed in o he ca ie bu e s eam ia a compu e -con olled injec ion al e. Resul s and Discussion Cul i a ion o Bacillus licheni o mis The cul i a ion o Bacillus licheni o mis was pe o med o p o ease p oduc ion. Ba ch and ed-ba ch cul i a ions in complex echnical media we e in es iga ed. The e men a ion media con ained hyd olyzed co n s a ch, soy meal, casein, and co n s eep liquo .The in o ma ion ob ained in he ba ch cul i a ions p o ided he basis o he ed-ba ch in es iga ions. In he beginning o he p ocess, biomass is p oduced in a pu e g ow h phase. The p o ease p oduc ion phase s a s when he glucose concen a ion alls below a ce ain concen a ion and he mic oo ganisms exc e e p o eases in o he medium o assimila e p o ein. Suc ose o mal ose can be added du ing he p oduc ion phase. This subs a e eed mus be con olled o op imal p o ease yield wi hou any lag phase. I he suga concen a ion is oo high, biomass will be p oduced and he p o ease p oduc ion will dec ease. On he o he hand, p o ease p oduc ion will dec ease when he subs a e concen a ion is below a c i ical le el ha esul s in dec easing cell ac i i y. Figu e 3 shows he simul aneous on-line analysis o mal ose, suc ose and glucose wi h he ou -channel enzyme he mis o o a ba ch cul i a ion o Bacillus licheni o mis. A compa ison o he on- line da a wi h di e en o -line me hods was p esen ed be o e (15). An in e es ing ela ion be ween he suga da a and he oxygen ans e A Fou -Channel Enzyme The mis o Sys em o Biop ocess Con ol 325 aD T T T! 3 | | - | & Pu a Ba I ||| ES | A: pa ed = | | 1 | ane = a 5 35[628 P ® 3 742 2 3 15 E D> . 2 o 3 a 1 10 2 iq 45 5 { | 8 0 05 10 1 3 e men a ion ime (h) Fig. 3 Glucose( o)-, mal ose( ¢ )- und suc ose concen a ion (=) du ing a ba ch cul i a ion o B. licheni o mis compa ed wi h he oxygen ans e a e (OTR). a e (OTR) can be obse ed. In a i s phase he suc ose consump ion was no co ela ed wi h a change in he OTR amoun . A e he comple e consump ion o suc ose, he consump ion o glucose wi h a linea inc ease o he OTR s a . A e a lag phase caused by he comple e glucose consump ion, he OTR da a inc ease is longe . A ed- ba ch cul i a ion moni o ed wi h he ou -channel enzyme he mis o is p esen ed in Figu e 4. In compa ison be ween mal ose concen a ion and cul u e luo escence signal demons a es an impo an di e ence be ween he ba ch and he ed-ba ch p ocess. In he ed-ba ch p ocess he cul u e luo escence signal inc eases ea lie , and eaches highe le el. The cul u e luo escence Signal is co ela ed wi h he amoun o li ing cells and hus is an indica o o ceil g ow h. The esul s o his S udy indica e ha mal ose as C-sou ce is esponsible o he cell g ow h, because he cul u e luo escene inc eases wi h he mal ose consump ion. A small peak in he cul u e luo escence signal can be obse ed a e he comple e consump ion o glucose, and indica es he change om glucose o mal ose consump ion. This peak appea s ea lie in he ed-ba ch p ocess, causing a smalle amoun o glucose in he s a medium and indica es he apid consump ion o glucose. Du ing ea lie expe imen s i has become ob ious ha he mul iple addi ion o mal ose has no posi i e e ec on he p oduc ion o p o eases. Howe e , a de ailed analysis wi h he enzyme he mis o was success ully pe o med o suc ose analysis. The esul s o his 326 H.-G. Hundeck e al. s udy indica ed ha he addi ion o suc ose should imp o e he ane | | ® oO o 5 | 3 @ | Ss = | © 3 3 = 55,68 5 2 8 20 8 }5 2 < 42 3 15 E mo e a: . +10 2 5 4 0 + - o0z 0S Ten: ed-ba ch-cul i a ion 20 3 5 3 iss [43 E a °, 0: I?» 2 5 b 4 "0 0 0 5 10 15 e men a ion ime (h) Fig. 4 Glucose( no )-, mal ose( © )- und suc ose concen a ion ( =) du ing a ba ch and a ed-ba ch cul i a ion o B. licheni o mis compa ed wi h he cul u e luo escence signal. p o ease p oduc ion. Two ed-ba ch e men a ions wi h con ol o suc ose addi ion wi h he ou -channel enzyme he mis o a e p esen- ed in Figu e 5. I he suc ose concen a ion is oo high, biomass will be p oduced while he p o ease p oduc ion dec eases. The p o ease yield was measu ed on-line by a s opped- low-FIA (16). The da a o he p o ease concen a ion we e i ed and di e en ia ed o ge he p o ease p oduc ion a e. Immobilized whole cells o Saccha omyces ce e isiae we e used ins ead o he enzyme columns o moni o he concen a ion o assimilable suga s du ing a cul i a ion. The cells we e immobilized in calcium algina e. The hea p oduced by he immobilized cells a e addi ion o di e en mono- and disaccha ides is shown in Figu e 6 as empe a u e change e sus concen a ion. The immobilized cells p oduce hea du ing he assimila ion o di e en suga s in he medium. Thus, his mic obial senso doesn’ de ec a single compound bu he A Fou -Channel Enzyme The mis o Sys em o Biop ocess Con ol 327 amoun o all assimilable subs ances in he medium. This pa ame e is ie 80 7 : 2 a © i «604 gh : oe ae io | & e - y } a Y w e ne % =; a ° ne 000° u ES FIRE Ea oa @) Sa e En 516 RR one 5) en nl : 10 20 a e men a ion ime (h) Fig. 5 Compa ison o on-line da a o suc ose de e mina ion using he enzyme he mis o wi h p o ease p oduc ion a e du ing wo ed- ba ch cul i a ions o B. licheni o mis. < 10 E © oe Ro} = 4 6 4 = © 4 glucose 7 44 * suc ose = J = uc ose = 23 mal ose 4 oF 2 3 4 concen a ion (g/l) Fig. 6 Hea e olu ion o di e en suga s measu ed wi h immobilized cells o S. ce e isiae. highly in e es ing o use in echnical media, in which he concen a ion dis ibu ion o di e en subs a es is o en nea ly unde ined. I would be p e e able o use he same o ganisms in he mic obial senso as in he cul i a ion o be moni o ed. Howe e , he use o yeas cells demons a es he p inciple and po en ial o his analy ical me hod. The 328 H.-G. Hundeck e al. da a o he mic obial senso du ing ba ch cul i a ions we e shown be o e (15). Enan ioselec i e analysis The p inciple o de ec ion o enan iome ic excess is based on wo enzyme eac ions (Figu e 7). One o he used enzymes (a- chymo ypsin) can only eac wi h one enan iome , p e e ably he L- s e eospeci ic compound. The o he used enzyme eac s wi h bo h s e eospeci ic compounds. Bo h enzymes eac wi h he sample. D,L-aminoacid es e + 0 ES RCOO’+H"+ ROH a-chymo ypsin L-aminoacid es e + H,O > RCOO’+ H'+ ROH Fig. 7 Enzyme eac ions o enan iome ic analysis The hea gene a ed du ing he eac ion is measu ed ia he enzyme he mis o . The concen a ion o bo h enan iome ic compounds was ob ained by his eac ion and he enan iome ic excess was calcula ed om his da a. The empe a u e signal inc eases in a linea way by a eac ion o a-chymo ypsin wi h di e en concen a ions o a 50:50 acema e o D,L-es e s, he eac ion wi h he es e ase esul s in a wice as high empe a u e signal (Figu e 8). Ss 40 E SC 5 B 304 © a © a wee § 3 104 04 - T 0 10 20 phenylalanineme hyles e (mM) Fig. 8 Calib a ion unc ion o eac ion o an aminoacides e 1.)L- (=) und D,L-es e s (+) wi h a-chymo ypsin 2.) L- (a) und D,L-es e s ( + ) wi h es e ase A Fou -Channel Enzyme The mis o Sys em o Biop ocess Con ol 329 By using he L-es e bo h enzyme eac ions gi e he same empe a u e signal. By a ia ion o he D-compound in a mix u e o bo h enan iome es i could imp essi ely be shown ha he empe a u e signal o he a-chymo ypsin eac ion depends on he amoun o he D- compound and he empe a u e signal o he es e ase eac ion was cons an caused by he same concen a ion as a whole, o bo h compounds (Figu e 9). By combining he whole concen a ion o D,L- compounds, ob ained by he es e ase eac ion and he concen a ion o he L-compound, a ailable om he a-chymo ypsin eac ion, i is possible o calcula e he enan iome ic excess. This p ocedu e is a new p inciple o analyze enan iome ic mix u es. The same eac ion unc ioned well in solu ions ha con ained dime hyl o mamide. The esul s o his s udy indica ed ha enan iospeci ic de ec ion is possible also in o ganic sol en s. 40 y = 31,405 - 4,2857e-4x R*2 = 0,000 a a z 7 a y = 27,362 - 0,27457x R*2 = 0,996 empe a u e signal (mK) wo oO 1 10 J 0 20 40 60 80 100 ec (%) Fig. 9 Calib a ion unc ion o a eac ion o an aminoacides e acema e by a ia ion o he D-compond wi h a-Chymo ypsin ( « ) und Es e ase ( = ) Summa y A biosenso sys em was de eloped and sucess ully applied o Cul i a ion moni o ing in bio echnology. To achie e high eliabili y and long e m s abili y, he ou -channel enzyme he mis o sys em was in eg a ed in o a FIA sys em. The en i e biosenso sys em included cell- ee sampling, sample condi ioning, and analysis. The sys em was au oma ed and could be un as a s and-alone e sion o on-line moni o ing. The sys em o e s he possibili y o ob ain a mo e de ailed insigh in o he whole biop ocesses o Spodop e a ugipe da and Bacillus licheni o mis and can se e as a basis o p ocess con ol.